LM-Kit.NET

Embeds on-device AI in .NET applications: document extraction, OCR, retrieval with citations, text analysis, speech and agents run in-process with no required cloud calls.

Suggest a correction

Overview

LM-Kit.NET is a NuGet package that runs language, vision, embedding and speech models inside a C#, F# or VB.NET process. Developers extract typed fields with per-field confidence from PDFs, Office files and scans, convert documents to Markdown with OCR, build retrieval and PDF chat that return the document, page and passage behind each answer, and run entity, PII, sentiment and classification analysis on text and images. Agents call built-in or MCP tools under permission policies with approval gates. Models run on CPU, CUDA, Vulkan or Metal on Windows, Linux and macOS; production use is free below published revenue, headcount and funding thresholds and otherwise needs a custom Professional license.
More product details

Product

Documentation
docs.lm-kit.com/lm-kit-net/index.html
License ID
  • LM-Kit End User License Agreement (https://lm-kit.com/legal/eula/)
License Model
Proprietary
Release Notes Url
docs.lm-kit.com/lm-kit-net/guides/changelog.html
Solution Kind
Platform

Who it is for

Industries
  • Healthcare
  • Financial services
  • Government
  • Manufacturing
Organization Types
  • Business
  • Nonprofit
  • Government
  • Education
  • Individual
Roles
  • Developer
Use Cases
  • Information extraction
  • Document question answering
  • Invoice data extraction
  • Document summarization

Features

Text Analysis

Sentiment and emotion analysis

Keyword extraction

More details & conditions: Text Analysis

Custom classification

Named entity recognition

PII extraction

Categories: Natural Language Processing

Sentiment Analysis

Positive, negative and neutral sentiment classification with multilingual support

Emotion detection (joy, anger, sadness, fear, surprise, disgust)

More details & conditions: Sentiment Analysis

Sarcasm detection

LoRA fine-tuning on domain data

Categories: Natural Language Processing

Machine Learning

Local model inference on CPU, CUDA, Vulkan and Metal

LoRA adapter training and loading

More details & conditions: Machine Learning

Text and image classifiers

Embeddings and reranking

Categories: Natural Language Processing

Data Extraction

Schema-driven structured field extraction with per-field confidence and human-review flag

Named entity recognition with character offsets

More details & conditions: Data Extraction

PII extraction with character offsets

Keyword extraction

Categories: Natural Language Processing

Natural Language Processing

Sentiment, keyword, entity and PII analysis

Text classification and language detection

More details & conditions: Natural Language Processing

Summarization, translation, rewriting and grammar correction

Speech-to-text transcription

Categories: Natural Language Processing

Topic Classification

Custom classification into user-defined categories on text and images

Best-category and ranked top-category results with confidence scores

More details & conditions: Topic Classification

Zero-shot document classification

Categories: Natural Language Processing

AI/Machine Learning

Local LLM, vision-language, embedding and speech model inference

In-process LoRA and full fine-tuning

More details & conditions: AI/Machine Learning

Model quantization

Embeddings and reranking

Categories: Natural Language Processing

Workflow details

Specialism
  • Embedded on-device AI runtime for .NET applications covering document intelligence, retrieval, text analysis, speech and agents

Categories: Other Artificial Intelligence AI

Workflow
  • Parse
  • Extract
  • Convert to Markdown
  • OCR
  • Split
  • Redact
  • PDF operations

Categories: Other Artificial Intelligence AI

Objects
  • Documents and scans
  • Extracted fields with confidence scores
  • Entities and PII mentions
  • Embeddings and vector stores
  • Agents and tools

Categories: Other Artificial Intelligence AI

Inputs Outputs
  • Name
    Documents and scans
    Description
    Accepts PDF, DOCX, XLSX, PPTX, HTML, Markdown, EML and images; returns typed JSON fields, Markdown, positioned elements and grounded answers
  • Name
    Audio
    Description
    Accepts audio files and streams; returns timestamped transcript segments

Categories: Other Artificial Intelligence AI

Inputs
  • Name
    Document and image inputs
    Description
    PDF, DOCX, XLSX, PPTX, HTML, Markdown, images and scans; OCR covers 34+ languages and speech-to-text covers 100+ languages

Categories: Natural Language Processing

Workflow boundary and exception handling
  • Each extracted field carries a confidence score and a human-review flag
  • Smart Redaction lets a reviewer approve each finding before permanent removal

Categories: Natural Language Processing

Pricing

FreeFree option

Free

Free

Commitment Months
0

Plan details & limits

Business Model
  • Free
Limits
  • Metric
    Free production-use eligibility
    Conditions
    Free for production use and redistribution when the company has under $1M USD annual gross revenue, 10 or fewer employees and no more than $3M USD raised from outside investors, assessed with affiliates; evaluation and development are free at any size, and personal, education, nonprofit and open-source use is always free. Free production use requires a visible acknowledgement of LM-Kit.

Professional

Contact vendorBilled annually

Contact vendor

Billed annually

Plan details & limits

Business Model
  • Subscription

Technical details

AI

AI usage conditions
Usage unit
Unmetered
Usage conditions
  • No per-token or per-seat fees; models run on customer hardware
Action Permissions
  • Action
    Run tools registered with an agent
    Approval
    Approval gates can be required per tool
    Permissions
    • Allow/deny patterns, maximum risk level and approval requirements set in a central policy
Citations
Yes
Customer Knowledge
  • Retrieval
  • Uploads
  • Connectors
Human Approval
  • Configurable
Input Modalities
  • Text
  • Image
  • Audio
  • Structured data
Output Modalities
  • Text
  • Structured data
Use Cases
  • Information extraction
  • Document question answering
  • Text summarization
  • Translation
  • Sentiment analysis
  • Text generation
  • Text improvement
  • Invoice data extraction

API

Access
Public
Programmatic access
Yes
Documentation
docs.lm-kit.com/lm-kit-net/api/LMKit.html

Deployment

Client OS
  • Platform
    Windows
  • Platform
    Linux (x64 and ARM64)
  • Platform
    macOS
Delivery Models
  • On premises
Isolation Modes
  • Air gapped
Offline Scope
  • Full workflow
Server OS
  • Platform
    Windows
  • Platform
    Linux (x64 and ARM64)
  • Platform
    macOS

Implementation

Delivery Mode
  • Self service

Data

Export Formats
  • Markdown
  • JSON
  • HTML
  • PDF
  • DOCX
Import Formats
  • PDF
  • DOCX
  • XLSX
  • PowerPoint presentation
  • HTML
  • Markdown
  • PNG
  • JPEG
  • TIFF
  • Plain text

Customization

Skill Level
  • Developer required

Reviews

User experiences, moderated before publication. Review policy

Write a review

No reviews yet.

Be the first to share your experience with LM-Kit.NET.

Add a product

Search to find a product.

Privacy choices

Choose what to allow.

You can change your choice here at any time. Refusing optional analytics does not limit the catalog.

AdvertisingAdSense is not currently enabled. Advertising choices will be available through its consent platform before ads are activated.Not active

Cookie details · Privacy rights